How to construct array for 45 degrees integral image

Hi…

I have an images M X N, and i would like to calculate rotated integral images (RSAT). Could anyone explains step by step to construct array for 45 degrees integral image .

Actually am attempting to implement rotated integral image, based on paper "An Extended Set of Haar- like Features for Rapid Object Detection" by Reiner Lienhart and Jochen Maydt. It mentions, that RSAT(x,y) can be calculate using 2 pass equation over a pixels.

I would like to know, should i rotate the image before i'm calculating RSAT or i just calculate RSAT using original image(without rotate it).

Thank you.

[Walter Roberson]

I am not sure what you mean by "integral image" ?

[rhn]

integral image mean rotated summed area table for 45º rotated features.

[Walter Roberson]

Perhaps you could post a small example or a link to a small example (that doesn't require that we read and understand the entire paper) ?

Answer 1

[Nitin]

you can look into imrotate to change rotate the image?

[rhn]

you mean that, first step I need to rotate the image , then calculate the rotated integral images?